Title :
A new generation robot system for object picking
Author :
MÜller, Markus ; WÖrn, Heinz
Author_Institution :
Inst. for Process Control & Robotics, Karlsruhe Univ., Germany
Abstract :
Robot systems for object picking from unsorted heaps of objects (ROs) have been investigated for more than twenty years. Nevertheless, the developed systems are not flexible and reliable enough for the widespread requirements of today. The paper presents first a classification of ROs according to the manner of sensory acquisition and processing of knowledge about the scene. Based thereupon, a new generation robot system is introduced, which involves reliable object picking without unintended lifting and damaging of neighbouring objects, whereas a minimum of previous knowledge about the scene is required. The robot system was realized and tested successfully in the shipping department of a German catalogue sales company.
Keywords :
inference mechanisms; materials handling; probability; robot vision; sensor fusion; uncertainty handling; German catalogue sales company; active sensing; new generation robot system; object picking; probabilistic reasoning; sensory acquisition; shipping department; unsorted heaps; world model; Artificial intelligence; Employment; Information processing; Intellectual property; Intelligent robots; Layout; Marketing and sales; Process control; Robot sensing systems; System testing;
Conference_Titel :
Multisensor Fusion and Integration for Intelligent Systems, 2001. MFI 2001. International Conference on
Print_ISBN :
3-00-008260-3
DOI :
10.1109/MFI.2001.1013508